Offshore wind is old enough now that the first farms are starting to wear out, and hardly anyone’s taken one apart yet. Denmark took down the 11 little turbines at Vindeby, the world’s first offshore wind farm, in 2017, and England cleared the two at Blyth in 2019. That’s pretty much the whole list of finished jobs. So the one sitting off the east coast of Ireland is interesting for a reason the industry would probably rather it wasn’t: it’s a 22-year-old wind farm that quit working, got half-dismantled, and is now waiting on paperwork to come down the rest of the way.<\/p>\n
If you’ve sailed past Arklow Bank in the last year, you’ve seen it. Seven steel towers, 406 feet to where the blade tips used to be, standing on a sandbank about six miles off County Wicklow with nothing turning on top of them. The blades came off in the summer of 2025. The towers, the nacelles (the boxes on top that hold the gearbox and generator), the monopiles under the water and roughly 14 miles of cable are all still out there.<\/p>\n
And according to the company that owns it, they’re getting more dangerous every year they stay.<\/p>\n
Arklow Bank Wind Park Phase 1 was a big deal when it went in. GE built it in 2003 and 2004 with Airtricity, an Irish developer, as a demonstration site for its then-new 3.6-megawatt offshore turbine, and GE’s own announcement<\/a> called it the first offshore wind project anywhere with turbines rated above 3 megawatts. Seven machines, 25.2 megawatts combined, each with a 341-foot rotor on a hub 241 feet above the sea. It started sending power to the Irish grid in 2005, and it’s still the only offshore wind farm in Ireland that ever has.<\/p>\n Then it got old. One of the seven was hit by lightning on October 19, 2022, during a thunderstorm, and burned until there wasn’t much left of its blades. The rest kept going until May 2024, when the last three turbines still running were switched off. The company’s 2024 accounts, as reported by RT\u00c9<\/a>, put that down to those three turbines “incurring major faults” in the second quarter, and say they were deactivated for safety reasons after reaching the end of their cycle.<\/p>\n Which, frankly, is about as much detail as anyone outside the company’s going to get. GE Vernova, the spun-off GE energy business that now owns the operating company, Arklow Energy Ltd, hasn’t said publicly what the faults were. A 20-year-old gearbox 241 feet up in salt air can go wrong in a lot of ways, and I wouldn’t want to guess which one.<\/p>\n The money side tells the same story in fewer words. Revenue fell 76 percent in 2024, from 3.66 million euros to 889,003 euros, and the company booked a loss of 14.2 million euros for the year, most of it a 13.7-million-euro provision for what it called a blade maintenance campaign. That campaign’s what you can see from the shore now.<\/p>\n You don’t send a ship, exactly. You send a barge that stops being a barge.<\/p>\n The vessel GE Vernova used was a jack-up called Typhoon 3000, which is basically a flat steel pontoon about 197 feet long and 92 feet wide with four legs at the corners, each 171 feet long. Towed into position, it drops those legs to the seabed and then winches its own hull up out of the water until it’s standing on them like a table. The waves keep moving. The deck doesn’t. That’s the only way you’ll get a crane steady enough to unbolt a blade more than 160 feet long from a hub 241 feet up without the hook swinging around. We covered a much bigger version of the same trick when a Belgian installation ship climbed out of the North Sea<\/a> to put turbines up, and this is the same idea run backwards.<\/p>\n Arklow Bank is a sandbank, so the water around the towers is only 10 to 23 feet deep, which you’d think makes life easier. It doesn’t. That’s too shallow for the big installation ships and exactly right for something like Typhoon 3000, which doesn’t have its own engines and had to be towed out by an anchor-handling tug. The marine analytics firm Spinergie<\/a>, which tracked the vessels through the job, says the barge carried a rack built to hold nine blades, which is three turbines’ worth. So the crew made two round trips out of the Port of Waterford, on Ireland’s southeast coast, and came back each time with nine blades laid in the rack.<\/p>\n The campaign ran from June 7 to July 13, 2025. Thirty-six days for the barge, with a median of 3.4 days parked at each tower and a couple of stops that stretched to six. Eighteen blades off six turbines, because the seventh had already lost its blades to the fire. The weather cooperated, with waves averaging under 3 feet, and the blades went back to Waterford, where they’re in storage.<\/p>\n Because blades are what’s most likely to come down on their own. A fiberglass blade that’s been pitching in Irish Sea wind for two decades with no one maintaining it is a liability in a way a steel tube isn’t, at least not yet. Spinergie calls the 2025 job a “make safe” campaign rather than a cleanup, and I think that’s the right read: you pull the thing that could fall off and hit a fishing boat, and you leave the rest standing while the lawyers work.<\/p>\n And the lawyers have been at it for a while. Arklow Energy asked the Irish planning body, An Coimisi\u00fan Plean\u00e1la, to open pre-application talks in January 2025, sat through three meetings with it between June 2025 and March 2026, and got the consultation closed out in May. Then, in September, it filed the real thing: an application to dismantle and remove all seven turbines, the foundations, an onshore substation, 4.6 miles of inter-array cable and 9.7 miles of export cable, all of which has to be dug out of the seabed, cut and hauled away.<\/p>\n The cover letter is where it gets blunt. Jerry Barnes, a director at the planning consultancy MacCabe Durney Barnes, wrote on the company’s behalf that the degrading structures pose “an increasing risk to the safety of the maritime area,” per the Irish Times<\/a>, and warned of falling debris that could wash up on the coast. The paper also reports that a preliminary engineering review found the degradation will increase exponentially beyond 2027, which isn’t the kind of phrase you put in a filing unless you want the regulator to move.<\/p>\n How much steel are we talking about? GE’s own fact sheet from when the farm went up listed each complete turbine at 290 metric tons, roughly 320 short tons, with the tubular tower at about 160 metric tons (176 short tons) and each monopile at 280 metric tons (309 short tons), driven into the sand with a 16-foot footprint. Multiply by seven and you’ve got somewhere north of 4,000 short tons of hardware sitting in 20 feet of water, minus the blades. Arklow Energy hasn’t published a cost for getting it out, and I couldn’t find one anywhere else, so we’re guessing.<\/p>\n Here’s the catch. In its December 2025 directors’ report, the company said it expected to start decommissioning in the summer of 2027. By the time the September filing went in, that’d slipped: Barnes told the commission that if consent arrives in the first half of 2027, the offshore work could happen in 2028. So the towers get at least two more winters, and the company is openly asking the commission to hurry.<\/p>\n Will it? Arguably the commission has bigger fish on its desk. Five large offshore wind farms in the Irish Sea are in the planning system right now, including an 800-megawatt one that SSE wants to build on the very same sandbank, with turbines up to about 940 feet tall. The old seven are, in a sense, in the way of the new 47 to 56, and I’d bet that’s not lost on anyone in Dublin.<\/p>\n Which brings up a question I don’t think anyone has answered publicly: will the monopiles actually come out? The meeting records from the pre-application talks say partial removal of the foundations was the likely approach, meaning the piles get cut off at some depth below the seabed rather than pulled whole. The records don’t say how deep. On a sandbank that shifts, that’s probably the sensible option. It also means Ireland’s first offshore wind farm will leave seven steel stumps under the sand more or less forever.<\/p>\nSo how do you get blades off a dead turbine six miles out?<\/h2>\n
Why only the blades?<\/h2>\n
The timeline is tighter than it looks<\/h2>\n